Having developed an interest in Neurology from having a child with ADD/ADHD, I found the program on PBS last week on “Brain Plasticity” fascinating. The key takeaway points are:
- The brain is much more dynamic in structure than we used to think. The brain molds itself (see factors below) not only when we are children, but the new findings are that this is a never ending process that can and does occur at any age
- Key factors driving brain plasticity are:
- Changes in stimulation ( loss of eyesight for example ). A fascinating study mentioned in the show is that when eyesight is lost, the 1/3 of the frontal cortex normally allocated for visual processing reallocates itself to other areas ( with auditory and touch sensory processing getting some of that space )
- Stimulation. Use it or lose it. Brain synapses formed previously will degrade over time if not “pinged” / touched by activity. Learning stimulates brain reconfiguration at any age
- Focus. The amount of reconfiguration/plasticity occurring is directly correlated to the attention/focus on the learning ( an interesting intersect with ADD/ADHD here. Note ADD is not Attention “Deficit” Disorder. It is Attention “Difference” Disorder. Those with ADD are apt to be less focused at times, but also have the ability to hyper focus ( think of a bell curve with a greater focus standard deviation for the population with ADD – imagine the focus abilities (and thus the learning capabilities if harnessed) on that right hand tail of the curve! )
- Challenge. What is the combination of “new task”, “level of difficulty”, and “level of interest”. The more a task is new, greater in difficulty, and viewed as interesting, the greater the learning and the greater the impact on the brain
Yes, there are also chemical actions at play related to brain health and brain plasticity. But here, I am focusing on learning and focus, and the rewiring process of the brain itself assuming no chemical level changes at play.
Speaking of rewiring – it has been thought that some of this rewiring process occurs during sleep ( converting short term to long term, and reconfiguring the long term storage for more efficiency ). I myself have the following postulations regarding the brain based on what I’v learned from neurologists, reading, observations of my son, and self observations:
- Memory of the human brain is virtually limitless. Look at Kim Peek ( the real “Rainman” ) – the person Dustin Hoffman was playing in the movie Rainman. He can continue to recall facts / figures / … from material read 20 years ago, or read 10 minutes ago. Zipcodes, phone numbers, dates, Presidents, biblical passages, and has shown expanded capacity over time to verbally express himself better ( interesting! ), and also has shown an equal skill at memorizing / recalling music in more recent years. Given a tune snippet, the writer, date, and a litany of other facts about the song are expressed by Kim. He does not have facts “rolling off the edge of the memory cliff” as he takes in new data by reading up to a dozen new books of information a day ( he reads the average page in less than 15 seconds and has 98% recall of information on the page )
- Cataloging of data and Filtering are at odds with memory recall. Kim has no filters due the frontal lobe not filtering the information due to a severe deficit in the portion of the brain that links the two larger brain halves together (the network bus between the two brain halves so to speak). What I’d like to know – is the cataloging of data ( assuming there is a cataloging process ) different for him or the same as the rest of us? If the same – we are all very efficient catalogers!
- Data is catalogued appropriately into long term memory when we sleep
- There is some type of meta data ( abstract cataloging ) structure applied that helps associate facts together ( perhaps responsibility for the content and location of the memory metadata is owned by the frontal cortex ) – an index of an index so to speak that the filtering leverages. For example, you know the person’s name begins with “B”, but you can’t think of it. Or you can think of three other moviestars of a movie and can picture the face of the main star given the movie name, but unable to retrieve the name of the main star. Sufficient retrieved catalog hits to form a useful (but incomplete) context, but not complete enough to solve the exact answer being sought
- Cataloging occurs during sleep – organizing memory using constructs something similar to a “Hashtable” or “Hashmap” you would find in a programming language Collection API. A memory may hit many buckets and 1 to n keys within a given bucket, each of which can point to a chain of 1 to N synapse connections, the chain ending in either an end marker, or another redirection/hub, the process repeating itself until end marker(s) are found when retrieved.
- Cataloging can be ( but may not always have to be ) very efficient. Much like what I’m describing above regarding science mimicing nature, perhaps there are different access speeds of reaching certain synapses ( due to filters most likely ). Kim Peek (who has no filters ) seems to have similar recall efficiency no matter how long ago or how frequently accessed the fact/material is ) depending on time since last accessed or frequency of access. Brain imaging has shown “hardening” of pathways in the brain based on the frequency of access ( much like a car and a field, and a dirt road forming in the field based on how many times the car has passed over the same path ).
- Processing occurs during sleep to move identified short term memory into long term memory, sometimes resulting in the actions of recataloging
- Processing occurs during sleep that “freshens” ( much like a keep alive thread ) memory synapse connections by “running the trail” – starting at one Hashtable bucket, finding the pointer (perhaps a region/table tuple), then running the chain of 1 to N synapses representing a portion of the memory until reaching the end (marker unknown), or another Hashtable is found ( repeating the process as necessary ). It is my contention that this “freshening” occurs in lighter sleep ( and deep sleep if no cataloging is required ). However, this “freshening” process is insufficient in and of itself to maintain memory integrity and brain health on its own
- Cataloging, and re-cataloging ( similar to a Hashmap split perhaps ) occurs while sleeping – deep sleep. Have you ever had a dream about an era long ago, and if the dream was remembered at waking, trying to recall other facts around the era covered in the dream? I have found near crystal clear recall of facts “related to” the dream ( but not in the dream per-se ). Lets say I had a dream with the dream (at least at first) occurring in a given era/timeperiod. Upon wake-up, I find substantially more clear and faster recall of items that are in or near the timeframe era of the dream (but not in the dream). Friends names in gradeschool, who the neighbors were, what the yard looked like ( detailed level ) at the residence lived at during the era in question. Upon wake-up, I can have that ”as if it happened yesterday” clarity of recall on the primary era of the dream ( if there was a primary era ).
- Cataloging forms new Hashmaps of related ( by what? ) data and those new ”Hashmaps” of data are indexed by a Hashmap of Metadata
- Recataloging moves Hashmaps of data from one to another – better and more efficiently organizing the data. Each bucket in the map is visited and run to some length and then moved/reconfigured elsewhere in the brain. This would explain the “as it if happened yesterday” experience. Because it did ( at least from a storage location and the synapse formation representing that fact/memory standpoint )
- Recataloging can move 1 to N memory items – with each chain of a relocated Hashtable being followed to end of marker. Because of the multiple “threads” being followed, any dreams recalled from this activity will appear more random and choppier
Now an interesting question. With science modeling nature – what features of Oracle Coherence are common with the Brain? Or asked another way, what features of the brain are common with Oracle Coherence?
1) Are memory synapse “catalogs” duplicated / backedup in another location of the brain? Those suffering a lesion or damage to a particular part of the brain seem to have knowledge of information/relevance that experts think they shouldnt have. A server instance with Coherence can fail, and yet no information is lost – all data in memory/cache is still accessible. An interesting question.
2) Do the number of “synapse hops” impact recall speed much like the speed of a Coherence data read and the number of network hops it takes to reach the “cache” (catalog) of record. Which is faster?
3) What kind of “Key” structures does the brain have in it’s in memory synapse network cache? What role does time/era or age of original memory play? What kind of key structure is associated with the Metadata?
4) Do the catalogs representing the in memory synapse network in the brain have a “backing store”? Scientists I am pretty sure would say no today. However, science also said in the past that the Universe is not expanding and that water did not exist elsewhere in the solar system. And this may sound strange – a local or remote backing store?
5) Does the brain have the concept of listeners and generating events on when “data changes of interest” occur?
If this has piqued your interest on Oracle Coherence, visit here. It is as easy as working with the Java Standard Edition (J2SE) Collection library and provides a solution for critical and/or enterprise applications requiring a reliable / available / scaleable / high performing / minimal data latency Grid capable distributed in memory storage cache network.
Here is an opportunity for some focused learning to challenge your brain ( and help your Enterprise Architecture at the same time ):
How to get started with Oracle Coherence.
